TROLLingdefining,,building,,and,operating,an,open,archive,for,linguistic,data

Helene%N.%Andreassen,%PhDUiT%The%Arctic%University%of%Norway

2nd Workshop+on+Standards+for+Data+Citation+and+Attribution+in+Linguistics

University+of+Texas+at+Austin,+April+8B10,+2016

Outline

A"variety"of"competencies"and"skills

Cooperation

IdeaAmbition

Operationalservice

UiT,and,open,access

UiT+will+be+recognized+by+a+culture+for+active+dissemination+through+open%channels for%publishing,+as+well+as+through+exhibitions,+journals+and+the+media.

UiT%strategic%plan%2020

Digital,– above,all!

Main,ambition,Explore%and%develop%the%digital%possibilities,%and%use%these%to%strengthen%our%services%to%employees%and%students

Strategy,Take%a%central%position%in%the%work%with%archiving%and%dissemination%of%research%data,%locally%and%nationally

University%Library%strategic%plan%2020

Photo:&Rune&Ytreberg

How,TROLLing, came,to,be

1. Inquiry%in%2013%from%the%UiT%linguistics%community%turns%the%library’s%ambition%to%work%on%open%research%data%into%action

2. Working%group%put%together,%consisting%of%researchers%and%subject%specialists%in%linguistics,%OA%specialists%and%system%developers

3. Establishment%of%a%threeNmember%scientific%advisory%board

4. Development%guided%by%scientific%needs%and%international%solutions

5. Launch%in%June%2014

Photos:&flickr.com/photos/kimgskytte

TROLLingThe,Tromsø,Repository,of,Language,and,Linguistics

Archive,for,open,linguistic,data,and,statistical,code

• International%service,%open%to%researchers%across%the%world%for%upload%and%download

• Maintained%and%curated%by%the%University%Library

– Relevance%of%uploaded%data– Quality%and%comprehensiveness%of%metadata

– Description%and%format%of%uploaded%data

opendata.uit.no

The,platform

Guiding+principle:+Be+futureBoriented,+and+think+bigger!

TROLLing%built%on%the%Dataverse platform%(https://dataverse.harvard.edu/)• Allows%adding%of%other%types%of%datasets%using%the%same%tools%and%templates• Facilitates%harvesting%of%data%by%international%services• Complies%with%DataCite (https://www.datacite.org/)%

Adaption, of,metadata,templateHow,to,optimize,retrieval,of,data

• Topic%specification– Field– TimeNdepth– Topic

– FreeNtext%keywords

• Description– Abstract– File%content

Setting,of,requirementsHow,to,optimize,reuse,of,data

• Description– In%template– In%readNme%file– In%data%file%(column%headings,%file%name,%etc.)

• Persistent%file%format– NonNproprietary– Open– Standard%character%encoding%(UTFN8)

Update, to,come

Dataverse Version,4

• Recently%released• Trolling%in%the%process%of%being%migrated

• Important%improvements– Richer%and%more%flexible%metadata%template– Tagging%on%file%level,%improving%the%search%function– Improved%metrics:%views,%downloads,%citations,%shares

Citing,the,dataBuiltKup,of,dataset,citation

• Persistent%identifier– Doi shortly%available

• Data%description– “Replication%data”%or%other

• Version%indicator– Previous%versions%accessible

Requirements,on,reuse,of,data

• Standard%license%selected:%CC0– Meet%the%potential%problem%of%attribution%stacking

• Citation%in%line%with%good%academic%practice– Use%the%reference%as%provided– (Add%subset%info%if%appropriate)

TROLLingThe,Tromsø,Repository,of,Language,and,Linguistics

Archive,for,open,linguistic,data,and,statistical,code

• International%service,%open%to%researchers%across%the%world%for%upload%and%download

• Maintained%and%curated%by%the%University%Library%at%UiT%

• Assisting%and%educating%the%users– User%guides– Instruction%videos

! Blog%interface%for%communication! Cooperation%with%faculty

site.uit.no/trolling

TROLLingThe,Tromsø,Repository,of,Language,and,Linguistics

Archive,for,open,linguistic,data,and,statistical,code

• International%service,%open%to%researchers%across%the%world%for%upload%and%download

• Maintained%and%curated%by%the%University%Library%at%UiT

• Development%of%user%guides%and%promotional%material%in%cooperation%with%faculty

• Marketing%in%every%channel%possible– Promotion%material– YouTube

! Cooperation%with%faculty,%graphic%designers%and%video%producers https://www.youtube.com/watch?v=uEf0c0NT9_A

OutreachConferences,and,meetings:,presentations,and,workshops

Laura,Janda• Slavic+Cognitive+Linguistics+Conference,%U.%of%Sheffield%and%Oxford,%2015.• 13th International+Cognitive+Linguistics+Conference,%Northumbria,%2015.%• Palatalisation Workshop,%CASTL/UiT,%2014.

Helene,N.,Andreassen• Journées FLOraLBPFC:+PFC+dans+le+champ phonologique,%Paris,%2015.• Journées FLOraL (Français Langue ORAle et+Linguistique),%Paris,%2014.

Philipp,Conzett &,Leif,Longva• emtacl15+B emerging+technologies+in+academic+libraries,%Trondheim,%2015.

Philipp,Conzett &,Obiajulu Odu• Dataverse Community+Meeting,%Harvard,%Cambridge,%MA,%2015.

OutreachApproaching,the,publishers

• Encouraging%from%above– Journal%editorial%boards

• put%TROLLing%into%guidelines

– Cristin (Norwegian%National%Research%Information%System)• create%a%category%“data%collection”%or%“dataset”/make%it%count

• Encouraging%from%below– Networks%(TROLLing%team%and%UiT%linguists)

– UiT%based%journals– OJSNDataverse plugin%(TBT)– Individual%projects

OutreachVisibility,in,social,media

Facebook– where%“everything%now%happens”

• Updates– New%uploads– Presentations/workshops– Technical%information

• Collaborative%management– TROLLing%curators– Faculty%research%assistant

User,activity,in,TROLLing, (per,April,7,,2016)

Numbers• 40%studies• 1394%downloads• 105%registered%users

– 19%countries– Europe,%Asia,%NorthN and%SouthNAmerica

Contributors• 24%unique%contributors

– 5%countries– Europe,%NorthNAmerica

Associated,publications• OA%journals• Paid%journals• No%publication

• PhD%thesis• Master%thesis

User,activity,in,TROLLing, (per,April,7,,2016)

Content

• Subfields– Semantics,%syntax,%morphology,%phonology,%phonetics

– Synchronic,%diachronic,%first%and%second%language%acquisition

• Languages– Czech,%Old%Church%Slavonic,%Russian,%Ukrainian

– French,%Romanian,%Spanish– German,%Norwegian– Saami

Content

• Types%of%data– Tables,%charts– Audio,%video– Scripts,%experimental%method

Fully,operational,service,and,why,curation,is,still,necessary

TROLLing,identity,,a,clear,definition,It%is%a%place%for%open,%structured%datasets%belonging%to%the%science%of%langue

• Yes– Structured,%well%described,%openly%accessible%datasets

• No– Metadata%only– Primary%data– Sensitive%data

– Bibliographies,%dictionaries,%national%anthems

– (To%be%continued)

BUT• Researchers%have%little%time• Researchers%are%not%used%to%

think%about%data%management

VIA,CURATION! Assistance%and%training! Consistent%optimization

«Guidelines& so&easy&that&my&grandmother&would&not&have&any&problems&uploading&data.»

Member&TROLLing&Scientific&Advisory&Board

To,learn,more,about,TROLLing

• Visit%the%archive%at%opendata.uit.no

• Visit%the%blog%at%site.uit.no/trolling%

• Contact%us%at%trolling@ub.uit.no

• New,idea,Join%us%in%a%TROLLing%webinar,%where%we%can%have%a%look%at%the%archive%together,%live,%all%while%being%located%in%different%parts%of%the%world

Thank,you,for,your,attention*

Helene,helene.n.andreassen@uit.no

TROLLingtrolling@ub.uit.no

Tromsø,"April"3,"2016"(private"photo)

*Thanks&to&Philipp&Conzett,&Stein&Høydalsvik,&Laura&Janda,&and&Leif&Longva (UiT)&for&useful&information&and&constructive&comments